The Radar Target Recognition Research Based on Improved Neural Network Algorithm
Duan Peipei, Hui Li · 2014
High range resolution profile(HRRP) has been a hotspot in the field of radar target recognition. But there are still many problems. This paper focuses on one-dimensional range target recognition based on an improved BP algorithm. There are a large number of samples used to learn the distribution characteristics of radar targets. The BP neural network algorithm is applied to match the template. Though standard BP algorithm has many advantages, it sometimes can not even be trained. Here, the influence of activation function is analyzed and a new algorithm is improved based on the biological learning process. The simulations show that the new method is feasible and can save the training time of the neural network. On the other hand, the generalization ability of this revised algorithm is improved too.